Changing Evaluations


I think a player constantly improves his understanding of chess with experience.  --  Yasser Seirawan

A couple of months ago, I started working on my first Stokyo exercise to try and improve my ability to evaluate positions, come up with good plans, and analyze lines. If you want to read up on that first attempt, you can see my first blog post here (in three parts).

Shortly after doing that I was speaking to another player at a tournament and I mentioned I was looking for some good sources of positions to use for additional exercises. He said he had used Reassess Your Chess Workbook by Jeremy Silman in a similar manner and thought it would be a good choice. I already owned the book, though I hadn't went through it at all, so decided to take a look.

This past week I did the first problem position, one from an opening. Before starting on the exercise, I decided I would change up the process a little, in an attempt to verbalize my thoughts more than I did in my previous attempt. This is still a work in progress but I thought I would share my results again.

I spent about an 1 ½ hours on this position and approximately another hour on the write-up. I decided to change my methods in doing this, thus I’m calling it a mini-Stokyo. Black to move.


  • Black’s pieces are less mobile/developed.  The knight on e8 has very little scope and stands worse than either of white’s knights. The black knight on c5 is well developed but is being attacked by white’s DSB.
  • Black’s DSB has a little less scope than white’s.  Black’s LSB is still at its home square and has limited scope; however, it is attacking the pawn on f5.
  • Black’s rooks are not connected, though the f8 rook has pressure down the half-open file. White’s rooks stand a little better, being connected, with the rook on the d-file giving support to the d5 pawn (may play a role if white can remove black's d6 pawn).
  • Black’s queen is less active than white’s.


  • Material is fairly even. White is currently up a pawn but the white pawn on f5 is threatened.


  • Black’s king is more open and if the white pawn on f5 is not taken care of, white can put additional pressure on the black king, with the potential  of destroying black’s castled position.
  • White can potentially create a passed pawn on the d-file by trading the DSB for the knight on c5 if black allows it.
  • Due to the pawn on f5, black has a potential upcoming attack on the queen along the b1-h7 diagonal.
  • White has better development.

Pawn Structure:

  • Black’s pawn formation points to an attack on white’s kingside; white’s also points to an attack on the kingside.
  • White has doubled pawns on the f-file.

Plans for Black:

  • Needs to get pieces developed to better squares.
  • The pawn on f5 poses threats to the castled position.
  • There is a potential issue with the d5 pawn becoming a passer.
  • Regain material.


Based on all the above factors, I feel white stands slightly better (just a little better than equal). +=

The only real plan I see for black is to take on f5. The rest of the evaluation factors point to Bxf5 being the best move; developing a piece and creating an attack on white’s queen, allowing black a chance to improve the weaknesses on the board.


Here, I didn’t look at very many lines, as I was trying to get the basic evaluation down. However, the plan to develop, the threat of the pawn and f5, and being down in material point to taking on f5 being the first order of business.

1… Bxf5 2. Qc1. I didn’t care for Qb2 as I thought putting the queen on that diagonal gives options for the bishop on g7 to become more active. Probably isn’t much worse (still +=).

2… Nd3. This is the main line I looked at. 3. Bxd3 Bxd3 4. Re1 Nf6. Black gets the minors developed and will be easily be able to connect rooks. I think this line is equal =. 3. Qa1 Nf6 is fine for white, += with the same threat on c5.

2. Nde4 (or Nce4) Qh4 3. Bxc5 dxc5 +=. Though, I don’t think I like this line as much as I think the black queen is a little misplaced here.

2... Nxe4 3. Nxe4 Nf6 4. Nxf6+ Qxf6 5. Qc1 += (borderline =)

2. Bd3?? Bxd3 winning a rook and bishop for the LSB. 3. Qc1 Bxf1 4. Kxf1 -+ maybe followed by Nf4

Engine Eval

I used Stockfish 5 for checking. In the initial position, the eval is +=, with 1… Bxf5 2. Qc1 being the top suggestion. 2. Qb2 is a little worse, = at 26 ply.

One option for black, that I did not look at, was 1… gxf5. Stockfish evaluates the position being just as good as Bxf5. I did not look at this move as it didn’t meet some of the other considerations in the evaluation phase (improving development/mobility) and seemed to open the king up even more (decreasing king safety).

2… Nd3 3. Bxd3 Bxd3 4. Re1 Nf6 +=. The alternate 3. Qa1 Nf6 is =.

2. Nde4 (Nce4) Qh4 3. Bxc5 dxc5 +=

2… Nxe4 3. Nxe4 Nf6 4. Nxf6+ Qxf6 5. Qc1 is =. The better line was 3… Qh4 =, followed by f3 or Bd3. I had looked at those lines, following up with Bh6 but didn’t care for the resulting positions and didn’t evaluate them.

2. Bd3, is obviously a blunder. The line I looked there is still -+.

Silman Comparison

I didn’t look at the hint/flavor text before doing this exercise. In future attempts, I think I will continue to not look at that. In this case, it was just a comment that white had captured the pawn, exf5 in response to f5.

Silman doesn’t like Bxf5. After 2. Nde4, white has control of the e4 square, giving the line 2… Nf6 3. Bxc5 dxc5 4. Bd3, cementing control of that square. In my analysis, I didn’t like the lines that gave up the knight on c5 for white’s DSB. As above, Stockfish preferred 2. … Nxe4 in this line and evaluated it as =. Silman looked at that line but suggested 3. Nxe4 Bxe4 (?! ...Qh4 best) 4. Qxe4 Nf6 5. Qc2, calling it good for white, giving the bishop pair. Stockfish agrees that line is good for white (+=).

I agree about the bishop pair comment and don’t think trading down was good for black. Black needs to develop and keep some tension in the position. If white wants to trade down, then respond to it but since there isn’t any tactical weakness, keep options open.

One possible weakness, that I didn’t evaluate in that line, was the idea of white eventually pushing c4-c5, potentially cracking open the d-file.

His recommended move is 1… gxf5. Benefits of this move,

  • keeps control of e4
  • gives options of a future e4, activating the g7 bishop, or a possible f4 push
  • gives the open g-file to black, who wants to attack on the kingside anyway, with ideas of Kh8 and Rg8
  • space advantage on kingside

Missed Evaluations

Here is some things I missed, based on the recommendations in the problem solution from the book.

  • white as more queenside space due to pawns on c4 and d5
  • white eyeing more queenside space with a3 and b4 to chase the knight away and open lines for the rooks
  • black has more kingside space, though I did get the desire for a kingside attack correct.
  • e4 square controlled by white. I saw this just thought it wasn’t as big as a weakness as Silman did.
  • the move gxf5 creates a “misconception” of opening up the kingside. I can see how his idea has merit and how I should have looked more deeply into to plusses of the line.

Final Thoughts

Once I had identified the first move, I should have analysed all the lines that dealt with it. While I looked at gxf5 and Rxf5, I discarded them as not meeting my goals. I should have verbalized why I discarded them as options when I was looking at lines.

I’m less comfortable with evaluating space considerations and didn’t really think along those lines. I need to work on including thoughts on space in my evaluations, though I really need to read up more on that idea.

Finally, I looked at some potential weaknesses and plans for white, such as Bxc5, potentially breaking open the d-file but I didn’t evaluate other moves that might accomplish the same thing, focusing on the bishop instead.